Psilocybin-assisted mindfulness training modulates self-consciousness and brain default mode network connectivity with lasting effects

Smigielski and colleagues situate this study at the intersection of two approaches known to alter consciousness: serotonergic psychedelics (here, psilocybin) and contemplative meditation. Earlier research has shown that both practices can induce self-transcendent experiences and modulate large-scale brain networks, particularly the default mode network (DMN), which supports self-related cognition. However, it remains unclear which neural mechanisms underpin lasting psychological changes after psychedelic experiences and how contextual variables such as meditative training (set and setting) shape these drug effects. The introduction highlights evidence that meditation and psychedelics share phenomenological and neurodynamic features, and it frames the DMN—especially medial prefrontal cortex (mPFC) and posterior cingulate cortex (PCC) coupling—as a candidate substrate linking altered self-experience to therapeutic outcomes. The present study therefore aimed to test whether a single administration of psilocybin given within a 5-day silent mindfulness retreat would produce state-dependent and lasting changes in DMN functional connectivity, subjective ego-dissolution experiences, and psychosocial behaviour. Using a randomised, double-blind, placebo-controlled, between-subject design in experienced meditators, the investigators measured brain dynamics with resting-state and two meditation-state fMRI scans before and after the retreat, and assessed acute subjective effects and persisting psychological changes up to 4 months later. The authors framed this as the first imaging study to examine the neurodynamics of a potent psychedelic administered specifically within a mindfulness retreat setting.

This was a randomised, double-blind, placebo-controlled, between-subject study conducted in the context of a 5-day silent Zen-style meditation retreat (sesshin). Thirty-eight healthy, experienced meditators (23 males; mean age 51.66 ± 8.32 years) were randomised to receive either oral psilocybin (315 μg/kg; mean absolute dose 21.82 ± 3.7 mg) or placebo (lactose) on the fourth day of the retreat while participants continued the daily meditation routine. Two participants dropped out prior to the start of imaging, leaving final group sizes of 20 (psilocybin) and 18 (placebo). Written informed consent and ethical approvals were obtained, and study procedures followed the Declaration of Helsinki. Neuroimaging assessments were performed on the day before and the day after the retreat to capture pre-post intervention effects. Each fMRI session comprised three 7-minute eyes-closed blocks in a fixed order approximating a meditation progression: resting state (RS), focused attention (FA), and open awareness (OA). The order of these three measurements was randomly assigned across participants, and pre- and post-scan times were matched within 2 hours to reduce time-of-day confounds. Acute subjective effects were measured 360 minutes after administration with the Five-Dimensional Altered States of Consciousness scale (5D-ASC). Persisting effects were assessed at 4 months with the Persisting Effects Questionnaire (PEQ), from which a global positive-change score was computed. MRI data were collected on a 3T Philips Achieva scanner with high-resolution T1-weighted structural images and T2*-weighted echo-planar functional sequences (TR 2000 ms; TE 35 ms; 32 slices; slice thickness 4 mm). Preprocessing used SPM12 and included slice-time correction, realignment, segmentation, normalization to MNI space, and smoothing (6 mm FWHM). Motion and physiological noise were addressed using Artifact Detection Tools for outlier detection and CompCor to regress principal components from white matter and CSF, along with realignment parameters; the global signal was not regressed. Residuals were detrended and bandpass filtered (0.009–0.08 Hz). Functional connectivity analyses applied data-driven independent component analysis (ICA) using the Infomax algorithm in GIFT, with the optimal number of components estimated by a modified minimum description length criterion. Stability of components was assessed via ICASSO (20 runs); 39 components were obtained and automatically labelled against resting-state templates, yielding three DMN subdivisions of interest: posterior DMN (pDMN), antero-ventral DMN (avDMN), and antero-dorsal DMN (adDMN). For hypothesis-driven testing, a 10-mm sphere was centred on the ICA-derived peak in mPFC (MNI -6 44 -6) and three other DMN hubs (PCC and bilateral angular gyri) were defined anatomically using the Conn toolbox. ROI-to-ROI bivariate correlations were Fisher-transformed for individual-level analyses. Statistical analyses included repeated-measures two-way ANOVA for 5D-ASC (group × scales), two-sample t-tests for PEQ global positive-change scores, paired t-tests for pre-post imaging contrasts within groups, and two-sample t-tests on subject-wise post-minus-pre differences to compute group-by-time interactions. Imaging results were assessed at an FDR-corrected p < 0.05 threshold; additional Bonferroni corrections were applied where noted. Correlations (Pearson) linked acute OSB scores to connectivity changes and PEQ scores. A forward stepwise multiple regression tested whether changes in nine ROI-to-ROI connectivity metrics (three ROI pairs × three conditions) predicted PEQ outcomes in the psilocybin group.

Acute subjective effects: Psilocybin produced robust alterations on the 5D-ASC relative to placebo, indicated by a significant group-by-scale interaction (F 4,144 = 33.82, p < 0.0001) and main effects of scale and group. Post-hoc testing showed significant increases in Oceanic Self-Boundlessness (OSB) and Visionary Restructuralization (VisRe) (both p < 0.0001) and a smaller increase in Vigilance Reduction (VigRe; p < 0.05). No significant group differences emerged for Dread of Ego Dissolution (DED) or Auditory Alterations (AA). ICA and spatial maps: The ICA decomposition identified 39 components and three DMN subdivisions (pDMN, avDMN, adDMN). Paired t-tests and time-by-group interaction analyses across the three conditions revealed significant psilocybin-related effects localized to the anterior cingulate cortex (ACC)/mPFC within the antero-ventral DMN. These effects survived correction for multiple comparisons including an additional Bonferroni penalty. Specific reported contrasts include a pre-post increase in connectivity in the psilocybin group (paired t, ACC cluster: t = 5.90, k = 144 voxels) and a group-by-time interaction cluster including left caudate, mediofrontal orbital cortex, and ACC (t = 5.12, k = 1066 voxels). ROI-to-ROI analysis: Paired tests within the psilocybin group showed significant post-retreat decoupling during OA meditation between the ICA-derived mPFC ROI and three DMN targets: PCC (β = -0.16, t19 = -3.81, p = 0.003), right angular gyrus (β = -0.12, t19 = -2.42, p = 0.030), and left angular gyrus (β = -0.11, t19 = -2.34, p = 0.030). Group-by-time interaction tests under OA reached significance for mPFC–PCC (t36 = -2.65, β = -0.18, p = 0.019) and mPFC–left AG (t36 = -2.61, β = -0.17, p = 0.019), but not for mPFC–right AG (t36 = -1.47, β = -0.10, p = 0.149). All ROI p-values reported were FDR-corrected; after a conservative Bonferroni correction for three interaction tests the mPFC–PCC effect remained the more robust finding (corrected p reported as marginal for the set, while the mPFC–PCC paired t-test survived correction: corrected p = 0.009). A negative correlation was observed between OSB and the pre-post change in mPFC–PCC connectivity (r = -0.595, p = 0.006; corrected p = 0.018), indicating greater acute self-transcendence was associated with larger post-retreat decoupling during OA. Persisting effects and predictors: At 4 months the psilocybin group reported markedly higher global positive-change scores on the PEQ (mean 2.58, SD 1.32) than the placebo group (mean 0.65, SD 0.98; t36 = 5.05, p < 0.0001). Acute OSB scores during the drug session correlated positively with these follow-up changes (r = 0.66, p = 0.002). A forward stepwise regression in the psilocybin group tested nine Δ connectivity predictors and identified two significant predictors explaining 44% of variance in PEQ positive global change (F 2,17 = 6.79, p = 0.007): an increase in mPFC–PCC connectivity at rest predicted greater positive change (β = 0.41, p < 0.05), while a decrease in mPFC–right angular gyrus connectivity during FA also predicted greater positive change (β = -0.49, p < 0.05).

Smigielski and colleagues interpret the findings as evidence that psilocybin administered within a mindfulness retreat produces state-dependent and lasting modulations of self-referential brain networks. The most notable result was that psilocybin potentiated positively experienced ego-dissolution (OSB), and that higher OSB correlated with reduced post-retreat antero–posterior DMN coupling (mPFC–PCC) specifically during open-awareness meditation. This supports the notion that decreased coupling between anterior and posterior DMN hubs relates to transient loosening of self-boundaries and altered self-experience. The authors place these results within existing literature showing both meditation and psychedelics produce DMN modulations. They note prior reports of decreased mPFC–PCC coupling during acute psychedelic administration and consistent DMN deactivation during meditation. In contrast to the OA-specific decoupling, psilocybin increased anterior DMN (mPFC/ACC) connectivity at rest in this sample; the authors suggest this reflects increased dynamic repertoire or network flexibility, possibly mediated by 5-HT2A receptor-driven glutamatergic signalling that can alter local network integrity. They propose these dissociable effects across brain states—enhanced anterior DMN integration at rest but decreased antero–posterior and fronto-parietal connectivity during OA—may together facilitate adaptive self-referential processing and underlie lasting behavioural change. A key behavioural observation was that psilocybin participants reported significant positive changes at 4 months, and that both acute OSB and post-acute connectivity changes predicted these outcomes. The regression analysis indicated that higher mPFC–PCC connectivity at rest and lower mPFC–right angular gyrus connectivity during FA were associated with greater long-term positive change, which the authors interpret as evidence that increased context-dependent DMN flexibility may enable therapeutic or transformative processes. The discussion also considers mechanistic ideas such as "deafferentation" of parietal regions from prefrontal input as a correlate of mystical-type experiences and cites lesion-mapping and other studies in support of the IPL's role in self-transcendence. The investigators report no post-acute modulation of medial temporal lobe–DMN connectivity in their exploratory analyses, suggesting that different components of ego-dissolution may map onto distinct network changes. The authors acknowledge several limitations they consider important: the modest sample size, the exclusive enrolment of highly experienced meditators (mean ~5000 hours of practice) which restricts generalisability, and the specific retreat setting that may both attenuate adverse reactions and interact with drug effects. They caution against extrapolating to unprepared or clinically different populations and recommend careful screening and safety procedures. Finally, they conclude that this is the first report of psilocybin-induced functional connectivity changes in self-referential networks measured after a mindfulness retreat, and they argue the findings motivate further research on how mindfulness and psychedelics jointly influence brain dynamics and psychotherapeutic change processes.